Tweet
Posted on December 15, 2015 by Euan Mearns
Every year I have a bet with a good friend on where the Brent oil price will be in December the following year. Last year I estimated $56.50, my friend $99, so this year I am a clear winner, even though I’m out by about 49%. Brent is trading at $38 as I write. I can extract little satisfaction from this since I got “close” for the wrong reasons.
Let’s cut to the quick. My forecast for Brent at around this time next year in my BAU (business as usual) scenario is $37. This is grim reading for all those involved in and around the oil industry. Worse still, I think there is high probability that we see sub-$20 oil before the first quarter is out. But this is great news for consumers. The reason is gross over-supply sustained throughout 2016, helped by Iran coming back to full market with an additional 800,000 bpd.
In addition to BAU I present two other scenarios. Capitulation where OPEC throws in the towel and cuts 5 million bpd that sends the price back to $100. And Event where terrorist activities in Saudi Arabia (or elsewhere) sends the price towards $100. The world has 3 billion barrels in storage and this may hang over the market for years to come.
This article first appeared on the Energy Matters blog. Third parties are welcome to cross post but must leave this sentence and link in place.
BAU Demand Model
IEA quarterly supply and demand data are used as a basis for the forecasts. Here, production less oil sent to storage = demand, i.e. consumption. Demand goes through an annual cycle where Q3&4 are typically higher than Q1&2. The quarterly data typically define a linear trend rising at about 1.5% per annum. These linear regressions (Figures 1 and 2) are used to predict quarterly demand in 2016 (Figure 3). This assumes no major upset to the global economy and that economic malaise in many quarters is offset by low oil price stimulating demand in others.
Figure 1 IEA Q1 demand data 2010 to 2015.
Figure 2 IEA Q2 demand data 2010 to 2015.
Figure 3 Demand summary. 4Q 2015 is estimated. The forecast for 2016 is based on the quarterly regressions shown in Figures 1 and 2. 3Q and 4Q regressions not shown.
BAU Supply Model
Forecasting supply is more tricky. Throughout 2015, supply has been amazingly robust in the face of collapsing price. The global oil market has tremendous momentum built on the back of several years of $100 oil. This momentum is not easily switched off since new conventional oil projects may have a 5 to 10 year lead time. There will be dozens of new fields around the world at various stages of development that will continue to come on stream in the years ahead.
One consequence of this is that most of the major producers are maintaining production plateaus where new supply is sufficient to cancel natural declines. See for example Saudi Arabia, OPEC, Russia, Asia and the North Sea in the November edition of Vital Statistics. None of these areas are showing any sign of retreat and I anticipate this will continue through 2016.
The one region that is more responsive to price collapse is LTO (light tight oil) drilling in North America where the US rig count has collapsed from 1609 (10 October 2014) to 524 (11 December 2015). But even here production has been robust with only modest declines posted thus far. One reason for this is that there are still 524 rigs drilling. These are drilling better wells targeting sweet spots and for example LTO production in The Permian has continued to rise offsetting some of the declines elsewhere.
Figure 4 The supply forecast perhaps looks a bit simple but is based on N American LTO production going down 600,000 bpd offset by Iran going up 800,000 bpd while the rest of the world trends sideways.
The US oil rig count has resumed its plunge and this must leave its mark in 2016 although this will be partly offset by an enormous back log of drilled but uncompleted wells. I am guestimating that LTO production falls 600,000 bpd in 2016.
But the joker in the pack is Iran, widely expected to rejoin with full exports in early 2016 and I anticipate that production may rise by 800,000 by year end, a bit higher than IEA estimates. Hence my summary forecast sees a net rise of 200,000 bpd, to produce what is effectively a production plateau (Figure 4).
The resulting stock additions are shown in Figure 5. This picture is similar to that provided by the IEA. The forecast sees another 547 million barrels going to storage.
Figure 5 Deducting the demand model from the supply model produces this picture of stock change for 2016. The World is still awash in surplus oil by the end of the year.
The IEA in December’s OMR says:
New and spare storage capacity should be able to accommodate the projected extra 300 mb of stocks.
So I, and the IEA, see a situation where gross over-supply is maintained throughout 2016. It is difficult to see a price recovery with that backdrop. But how to convert these numbers into a price forecast? Past relationships between supply, demand and price is all I have to go on and I therefore use the same empirical model as used last year (Figure 6). This has many limitations, amongst other things it is grounded in the trends defined in the period Jan 2002 to May 2011. The relationship may since have changed. But this is all I have to go on. Setting the supply curve to 97 M bpd and demand to 96.1 M bpd provides an indicative price for Brent of $37 in December 2016, very close to today’s price.
Figure 6 Based on an original idea by Phil Hart. I’m not sure I fully understand the dynamic here any more, but an explanation is given here. Large time lags between price signal and market response cloud the picture. Oil supply was inelastic to price up to May 2011, but then supply responded and has been responding ever since.
However, in the interim, it is possible the oil price goes much lower since these things have a habit of being overdone. The last time we saw a crash like this was in 1999 and the price then went sub $20 in today’s money (sub $10 in money of the day). If history repeats, that’s where the market is heading. Demand will after all soften 1Q 2016.
I am not even going to attempt a forecast for 2017. Figure 5 suggests the market will return to balance in 2017 but already has a record 3 billion barrels in storage, equal to about 32 days of consumption. This stock volume will hang over the market for a long while after the fulcrum is crossed. And it is difficult at this point to predict how the US LTO industry will respond to the prospect of rising price. Will the industry be killed stone dead by the events of 2015 and 2016. Or will drillers return when the price begins to inch up?
With the World in turmoil there is of course no guarantee of BAU and two other scenarios are briefly outlined below.
Capitulation
The countries hardest hit by the ongoing price rout are in OPEC closely followed by Russia. The possibility remains that OPEC decides to abandon the market share policy and revert to protecting price. I estimate that to regain $100, about 5 M bpd would need to be withheld from the market. Remember that stocks will last 1000 days with stock draws of 3 M bpd.
I just don’t see either Saudi Arabia, Iraq, Iran or Russia agreeing to such massive production cuts and therefore give this option a very low probability.
Event
With conflict raging across the Middle East, it is somewhat miraculous that oil production has not been disrupted, apart from in Syria and Libya. Impoverished Saudi Arabia, Iraq and Algeria look increasingly vulnerable to insurgency. Destabilisation of Saudi Arabia in particular, leading to a Libyan style of civil war, could of course have a profound impact on oil markets.
Onshore fields, processing and export facilities in Saudi Arabia are particularly vulnerable, although well guarded. The Abqaiq processing facility, that processes around 5 M bpd was attacked by 4 al-Qaeda terrorists in February 2006. And a terrorist was shot in Abqaiq as recently as September 2015.
While an event such as this looks increasingly possible it is at the same time impossible to assign a probability. But even if 5 M bpd production was lost, that 3 billion barrel cushion of oil in storage would cover 600 days. The price may bounce to $100, sending the shale drillers back to work.
Concluding Comment
It has always been impossible to predict the oil price. The supply, demand and security dynamic today is as high as it can be. Anything can happen in the next 48 hours. But my estimate for December 2016 is that oil prices will be around where they are today having gone much lower in the intervening months. $37 causing much pain to ‘The Industry’ today will be seen as a relief in a year’s time.
Every year I have a bet with a good friend on where the Brent oil price will be in December the following year. Last year I estimated $56.50, my friend $99, so this year I am a clear winner, even though I’m out by about 49%. Brent is trading at $38 as I write. I can extract little satisfaction from this since I got “close” for the wrong reasons.
Let’s cut to the quick. My forecast for Brent at around this time next year in my BAU (business as usual) scenario is $37. This is grim reading for all those involved in and around the oil industry. Worse still, I think there is high probability that we see sub-$20 oil before the first quarter is out. But this is great news for consumers. The reason is gross over-supply sustained throughout 2016, helped by Iran coming back to full market with an additional 800,000 bpd.
In addition to BAU I present two other scenarios. Capitulation where OPEC throws in the towel and cuts 5 million bpd that sends the price back to $100. And Event where terrorist activities in Saudi Arabia (or elsewhere) sends the price towards $100. The world has 3 billion barrels in storage and this may hang over the market for years to come.
This article first appeared on the Energy Matters blog. Third parties are welcome to cross post but must leave this sentence and link in place.
BAU Demand Model
IEA quarterly supply and demand data are used as a basis for the forecasts. Here, production less oil sent to storage = demand, i.e. consumption. Demand goes through an annual cycle where Q3&4 are typically higher than Q1&2. The quarterly data typically define a linear trend rising at about 1.5% per annum. These linear regressions (Figures 1 and 2) are used to predict quarterly demand in 2016 (Figure 3). This assumes no major upset to the global economy and that economic malaise in many quarters is offset by low oil price stimulating demand in others.
Figure 1 IEA Q1 demand data 2010 to 2015.
Figure 2 IEA Q2 demand data 2010 to 2015.
Figure 3 Demand summary. 4Q 2015 is estimated. The forecast for 2016 is based on the quarterly regressions shown in Figures 1 and 2. 3Q and 4Q regressions not shown.
BAU Supply Model
Forecasting supply is more tricky. Throughout 2015, supply has been amazingly robust in the face of collapsing price. The global oil market has tremendous momentum built on the back of several years of $100 oil. This momentum is not easily switched off since new conventional oil projects may have a 5 to 10 year lead time. There will be dozens of new fields around the world at various stages of development that will continue to come on stream in the years ahead.
One consequence of this is that most of the major producers are maintaining production plateaus where new supply is sufficient to cancel natural declines. See for example Saudi Arabia, OPEC, Russia, Asia and the North Sea in the November edition of Vital Statistics. None of these areas are showing any sign of retreat and I anticipate this will continue through 2016.
The one region that is more responsive to price collapse is LTO (light tight oil) drilling in North America where the US rig count has collapsed from 1609 (10 October 2014) to 524 (11 December 2015). But even here production has been robust with only modest declines posted thus far. One reason for this is that there are still 524 rigs drilling. These are drilling better wells targeting sweet spots and for example LTO production in The Permian has continued to rise offsetting some of the declines elsewhere.
Figure 4 The supply forecast perhaps looks a bit simple but is based on N American LTO production going down 600,000 bpd offset by Iran going up 800,000 bpd while the rest of the world trends sideways.
The US oil rig count has resumed its plunge and this must leave its mark in 2016 although this will be partly offset by an enormous back log of drilled but uncompleted wells. I am guestimating that LTO production falls 600,000 bpd in 2016.
But the joker in the pack is Iran, widely expected to rejoin with full exports in early 2016 and I anticipate that production may rise by 800,000 by year end, a bit higher than IEA estimates. Hence my summary forecast sees a net rise of 200,000 bpd, to produce what is effectively a production plateau (Figure 4).
The resulting stock additions are shown in Figure 5. This picture is similar to that provided by the IEA. The forecast sees another 547 million barrels going to storage.
Figure 5 Deducting the demand model from the supply model produces this picture of stock change for 2016. The World is still awash in surplus oil by the end of the year.
The IEA in December’s OMR says:
New and spare storage capacity should be able to accommodate the projected extra 300 mb of stocks.
So I, and the IEA, see a situation where gross over-supply is maintained throughout 2016. It is difficult to see a price recovery with that backdrop. But how to convert these numbers into a price forecast? Past relationships between supply, demand and price is all I have to go on and I therefore use the same empirical model as used last year (Figure 6). This has many limitations, amongst other things it is grounded in the trends defined in the period Jan 2002 to May 2011. The relationship may since have changed. But this is all I have to go on. Setting the supply curve to 97 M bpd and demand to 96.1 M bpd provides an indicative price for Brent of $37 in December 2016, very close to today’s price.
Figure 6 Based on an original idea by Phil Hart. I’m not sure I fully understand the dynamic here any more, but an explanation is given here. Large time lags between price signal and market response cloud the picture. Oil supply was inelastic to price up to May 2011, but then supply responded and has been responding ever since.
However, in the interim, it is possible the oil price goes much lower since these things have a habit of being overdone. The last time we saw a crash like this was in 1999 and the price then went sub $20 in today’s money (sub $10 in money of the day). If history repeats, that’s where the market is heading. Demand will after all soften 1Q 2016.
I am not even going to attempt a forecast for 2017. Figure 5 suggests the market will return to balance in 2017 but already has a record 3 billion barrels in storage, equal to about 32 days of consumption. This stock volume will hang over the market for a long while after the fulcrum is crossed. And it is difficult at this point to predict how the US LTO industry will respond to the prospect of rising price. Will the industry be killed stone dead by the events of 2015 and 2016. Or will drillers return when the price begins to inch up?
With the World in turmoil there is of course no guarantee of BAU and two other scenarios are briefly outlined below.
Capitulation
The countries hardest hit by the ongoing price rout are in OPEC closely followed by Russia. The possibility remains that OPEC decides to abandon the market share policy and revert to protecting price. I estimate that to regain $100, about 5 M bpd would need to be withheld from the market. Remember that stocks will last 1000 days with stock draws of 3 M bpd.
I just don’t see either Saudi Arabia, Iraq, Iran or Russia agreeing to such massive production cuts and therefore give this option a very low probability.
Event
With conflict raging across the Middle East, it is somewhat miraculous that oil production has not been disrupted, apart from in Syria and Libya. Impoverished Saudi Arabia, Iraq and Algeria look increasingly vulnerable to insurgency. Destabilisation of Saudi Arabia in particular, leading to a Libyan style of civil war, could of course have a profound impact on oil markets.
Onshore fields, processing and export facilities in Saudi Arabia are particularly vulnerable, although well guarded. The Abqaiq processing facility, that processes around 5 M bpd was attacked by 4 al-Qaeda terrorists in February 2006. And a terrorist was shot in Abqaiq as recently as September 2015.
While an event such as this looks increasingly possible it is at the same time impossible to assign a probability. But even if 5 M bpd production was lost, that 3 billion barrel cushion of oil in storage would cover 600 days. The price may bounce to $100, sending the shale drillers back to work.
Concluding Comment
It has always been impossible to predict the oil price. The supply, demand and security dynamic today is as high as it can be. Anything can happen in the next 48 hours. But my estimate for December 2016 is that oil prices will be around where they are today having gone much lower in the intervening months. $37 causing much pain to ‘The Industry’ today will be seen as a relief in a year’s time.